This study investigated the temporospatial variation, chemical composition, and source resolution of fine particles (PM2.5) in the southeastern seas of the Taiwan Island. 24-hr PM2.5 was sampled simultaneously at two remote sites, the Green Island (West Pacific Ocean; WPO) and the Kenting Peninsula (northern Bashi Channel; BC), in four seasons. After sampling, the chemical fingerprints of PM2.5 were characterized and further applied to resolve the potential sources of PM2.5 and their contribution by using a receptor model on the basis of chemical mass balance (CMB), enrichment factor (EF), and backward trajectory simulation. It showed that PM2.5 concentrations in winter (10.8 μg m -3 ) and spring (12.0 μg m -3 ) (i.e., during the period of Asian Northeastern Monsoons; ANMs) were higher than those in summer (4.0 μg m -3 ) and fall (6.6 μg m -3 ). In terms of chemical composition of PM2.5, secondary inorganic aerosols (SIAs = NO3 -, SO4 2-, and NH4 + ) (56.7~67.2%) were the dominant component of water-soluble ions (WSIs) in PM2.5, while crustal elements (Mg, Al, Ca, Fe, and K) (44.0~61.2%) dominated the metallic contents in PM2.5. High EF values (> 10) showed that V, Mn, Ni, Cu, and Zn were potentially contributed from anthropogenic sources. Moreover, organic carbon (OC) (0.6 μg m -3 ) was superior to elemental carbon (EC) (0.3 μg m -3 ) in PM2.5. The OC/EC ratios higher than 2.0 showed the potential chemical formation of secondary organic aerosols (SOAs) in the atmosphere in winter and spring. Trajectory simulation indicated that high PM2.5 concentrations were mostly originated from North and Central China, Japan islands, and Korea Peninsula. Major sources of PM2.5 resolved by CMB receptor modeling were ordered as: sea salts (19.9%) > fugitive dust (19.8%) > industrial boilers (oil-fired) (10.8%) > secondary sulfate (9.8%) > mobile sources (8.0%).